Based on 7 days of per-container usage, with extra memory after OOM kills and room for CPU bursts. You get the YAML; you apply it.
Before · example container
Recommended requests
How a request is sized
Baseline usage, trend, and burstiness. Each adds its own headroom to the request.
For container worker
Baseline
01What does this workload normally need?
0.420 cores
Growth
02Is it trending up, down, or steady?
+5% headroom
Burstiness
03How spiky is its day-to-day?
+18% headroom
Recommended
0.520cores
Was
1.000
–48%
What you get
The YAML on one tab. Seven days of usage on the other.
Per-container YAML. Current and recommended, side by side.
Pick your aggregation. Avg through Max. 7, 14, or 30 days.
A profile per cluster. Production uses P99 for CPU and memory, non-production P50. You can override the percentiles.
Dismiss what doesn’t fit. It leaves your list.
- name: worker
resources:
requests:
cpu: 1000m
memory: 2Gi
limits:
cpu: 2000m
memory: 4Gi- name: worker
resources:
requests:
cpu: 320m # P99 × 1.05 × 1.18
memory: 780Mi # P99 × 1.04 × 1.22
limits:
cpu: 480m # request × 1.5
memory: 1.0Gi # max observed × 1.30line 4
1000m 320m
-680m
line 5
2Gi 780Mi
-1.2Gi
line 7
2000m 480m
-1.52
line 8
4Gi 1.0Gi
-3.0Gi
Trend
Steady
Volatility (CV)
0.18 · Low
OOM kills
0 in 7d
Restarts
0 in 7d
Four checks
OOM-aware
An out-of-memory kill restarts the pod, so the recommended memory limit comes from the highest usage observed, not an average.
Memory · 7 days
covers the peak
Two real spikes. An average misses the second. Our recommendation doesn’t.
After OOM kills
A container killed at its limit never shows how much memory it needed, so observed usage reads low. After kills, the recommendation raises memory until the kills stop.
7d · 3 kills detected
headroom restored
Kills cap what we can observe. Our recommendation pushes the limit higher until the cycle stops.
Bursty CPU
When peaks run far above normal load, we recommend removing the CPU limit so bursts are not throttled.
Bursty CPU · 7 days
recommended ceiling: none
Bursts are not throttled.
Replica-aware
The dollar figure on a recommendation is the per-pod saving multiplied by your replica count, including HPA-scaled workloads.
Replica-aware savings
$35.70 / pod × 40 pods
Total
$1428/ mo
Per-pod savings scaled by your replica count, including HPA workloads. The dollar number is the workload total, not a single pod.
The workflow
The YAML goes through the same review as any other change.
Step 01
One click. The snippet is on your clipboard.
Current · Recommended
Step 02
Paste into whatever you already use.
Helm · Kustomize · Terraform · ArgoCD · Flux
Step 03
Your CI, your review, your rollout. After the change lands, Kubeadapt waits out a cooldown and sizes the container again from new usage.
Cooldown set by profile · re-analysis
Install in 5 minutes
Free up to 100 vCPU. After that, $1.99 per vCPU per month, dropping to $0.60 at volume.